Frontiers in Psychological and Behavioral Science Apr. 2014, Vol. 3 Iss. 2, PP. 27-35 - 27 - Associations between Simulator Sickness and Visual Complexity of a Virtual Scene Iwan Kelaiah *1 , Manolya Kavakli 2 , Ken Cheng 3 1 Australian Centre of Field Robotics, The University of Sydney, NSW 2006, Australia 2 Department of Computing, Macquarie University, NSW 2109, Australia 3 Department of Biological Sciences, Macquarie University, NSW 2109, Australia *1 i.kelaiah@acfr.usyd.edu.au; 2 manolya.kavakli@mq.edu.au; 3 ken.cheng@mq.edu.au Abstract- This study examined the effects of simulator sickness (SS) as a function of the visual complexity of animated virtual actors (AVAs) and the virtual environment (VE) in a virtual scene. Visually complex stimuli may be attractive; however, studies on SS indicate the possibility of significant health risks outweighing the expected benefit in virtual reality (VR) simulations. This study used a series of simulations to teach the basic skills required for village fire fighting to manage fires caused by car accidents. The participants learnt in one of 4 experimental conditions; simple (simple AVAs and simple VE), simple world (lifelike AVAs and simple VE), simple AVAs (simple AVAs and lifelike VE) and lifelike (lifelike AVAs and lifelike VE). We predicted that: (1) SS ratings would increase with the scenes’ visual complexity and (2) simpler VEs would compensate for the effects of visually complex AVAs. Surprisingly, the results contradicted our predictions, with no effect of either variable. We discuss possible explanations for these results, and suggest future research directions to design safe VR simulations. Keywords- Animated-Virtual Actors; Virtual Environment; Simulator Sickness; Virtual Reality; Virtual Scene; Visual Complexity; Visualisation; Vection LIST OF ABBREVIATIONS 3D three-dimensional ANOVA analysis of variance ANTECATALYST animated Cataglyphis ants AVA Animated virtual actors FOV field of view GIF Graphic Interchange Format IQR interquartile range JPEG Joint Photographic Expert Group SS simulator sickness SSQ simulator sickness questionnaire VE virtual environment VEAF vehicle accident and fire VR virtual reality I. INTRODUCTION Developing visually complex scenes in virtual reality (VR) applications is becoming easier with advanced three dimensional (3D) modelling tools [1] and the latest 3D engines [2, 3] many of which are freely available. The application of digital content — animated virtual actors (AVAs) and the virtual environment (VE) — is often technologically driven or market-driven. In most cases, the technology is used as a showcase [4, 5]. In the domain of education and training with multimedia systems using VR simulations [6], there are two conflicting hypotheses: media-affect-learning and method-affect- learning. The media-affect-learning hypothesis states that using better technology when delivering learning content promotes learning. In contrast, the method-affect-learning hypothesis argues that, as long as the methods promote cognitive learning, the medium does not matter [7]. In examining both the media-affect-learning and method-affect-learning hypotheses, Moreno and Mayer [6] compared the influence of different levels of immersion in VR devices and personalised narration. The experimental results with college students indicated that there were no differences in learning outcomes between the use of a standard desktop and a head